How AI in Construction Is Changing Things for the Better
AI is ready for construction
Machine learning and artificial intelligence (AI) are no longer the reserve of science fiction novels. They're now rapidly redefining the world we live and work in.
And if there's one sector that's truly ready for disruption by technology, it has to be the construction industry.
Slow to change
Against a backdrop of relentless, fast-paced change elsewhere, the construction industry still relies on decades-old techniques. Just compare a vehicle manufactured in the 1920s to one today – and then do that for a construction project. Today's vehicles and their production processes are hardly comparable to those of the 1920s; the fundamentals of construction are much less well-evolved.
How to explain this industry-wide reticence to change? One explanation is a back-ward looking attitude being (fairly or unfairly) seen as prevalent within the industry.
The likely explanation is: as an industry forced to rely on razor-thin margins, it doesn't have the resources to invest in research and development. Project stakeholders have therefore continued to fall back on tried-and-tested methods which often hold the industry back – stunting productivity and creativity.
Signs of AI coming through
In 2016, research conducted by KPMG highlighted a growing number of construction professionals and project stakeholders using emerging technology to change the way the industry works.
At the same time, construction project stakeholders are seeing fast growth in data generation and collection. What we don't have is a way for this data and insight to smoothly pass from project to project. This would maximise efficiency, productivity and safety, and build a transparent, co-operative construction industry.
Enter machine learning!
What is machine learning – and what does it mean for construction?
One definition of machine learning is: 'an application of artificial intelligence that provides systems with the ability to automatically learn and improve from experience without being explicitly programmed'. Put simply, it involves computers learning to make decisions, instead of humans manually programming them.
Machine learning tools can analyse data – learning from previous assumptions – and use its insights to continuously improve its decisions. It then uses human intelligence to enhance its insights even more over time.
This is a real game-changer in the construction industry, particularly in the labour-intensive, time consuming design and planning stage. These processes often require months, if not years, of work and preparation, from a team of highly qualified experts. The work they do is often manual and monotonous. It might be based on fragmented information, leading to errors in both design and construction phases. Inefficient design, cost estimating and planning drive the industry's infamously base profit margins and levels of risk unacceptable in some industries.
The use of AI and machine learning will revolutionise every part of this process. Particularly in workflows requiring prediction, classification (e.g. identifying BIM elements in a model) and optimisation of processes with many criteria. Here are just three examples of how this will improve the construction industry:
- BIM classification: BIM elements underline a project’s bill of quantities, which form the basis for cost estimates. Instead of manually checking BIM elements, machine learning evaluates the topology, geometry, BIM element attributes and more to identify and classify a BIM element. It asks a human for those elements it is unsure of and learns that algorithm for the future, never needing to check that classification again.
- Knowledge engineering and management: the design and construction industries are highly dependent on specialist knowledge. Capturing and retaining this knowledge could transform the fortunes of the industry so that companies can train AI agents to automate rule-based workflows. This knowledge base can augment the machine learning methods where knowledge acquisition is costly and hard work (e.g. acquiring, cleansing and labelling data sets).
- Risk factors: rather than rely on guesswork, project managers can use machine learning to analyse huge amounts of data, not only to identify the most likely risk factors for a given project – but to accurately predict the impact those risks will have.
In short, every step of the construction planning process could benefit from machine intelligence. These AI algorithms will, by nature, become better over time at predicting the correct configurations to optimise cost and schedule – leading to to better bid pricing, one of the main drivers of the profitability in the industry.
Let experts be experts
Since machine learning is, by nature, robotic – it's relentless. It learns continuously, without fatigue or bias (unlike humans).
Far from diminishing the value of expertise, machine learning will fuel AI technology to give experts more time to offer up their true skills. With less time wasted on carrying out mundane tasks, experts will finally have the freedom to bring creativity to a project. They will be able to test assumptions and propose tweaks to make projects better, safer and more efficient.
It stands to reason that if there's more certainty and less risk involved in the design and planning stage, experts will be able to experiment, giving rise to original, innovative buildings. Creativity will be the winner.
The benefits of machine learning could even go beyond this. By increasing the productivity and improving the pricing accuracy, it could create a virtuous cycle in the industry. There will be more funding for research, development, continued optimisation in efficiency – and, ultimately, a radical transformation in the industry. Transformation that is long overdue!